Table 2.

Representative software tools predicting Sec substrates in Gram-negative bacteria.

Tool	Algorithms	Target	URL or reference
SignalP4	Artificial Neural Network (ANN)	Sec/SPI; Cleavage site	https://services.healthtech.dtu.dk/service.php?SignalP-4.1; [13]
SignalP5	Deep Neural Network (DNN)	Sec/SPI; Sec/SPII; Tat/SPI; Cleavage site	https://services.healthtech.dtu.dk/service.php?SignalP-5.0; [14]
Signal-BLAST	BLASTP	Sec/SPI	http://sigpep.services.came.sbg.ac.at/signalblast.html; [15]
Signal-3L 2.0	Hierarchical Mixture Model	Sec/SPI; TMH; Cleavage site	http://www.csbio.sjtu.edu.cn/bioinf/Signal-3L; [16]
PrediSi	Position Weight Matrix (PWM)	Sec/SPI	http://www.predisi.de; [17]
Signal-CF	Pseudo Amino Acid Composition;K Nearest Neighbor Classifier	Sec/SPI; Cleavage site	http://www.csbio.sjtu.edu.cn/bioinf/Signal-CF; [18]
LipoP	HMM	Sec/SPI; Sec/SPII; TMH	https://services.healthtech.dtu.dk/service.php?LipoP; [19]
SPEPlip	ANN; Regular Expression Search	Sec/SPI; Lipoprotein; Cleavage site	http://gpcr.biocomp.unibo.it/cgi/predictors/spep/pred_spepcgi.cgi; [20]
Phobius/ PolyPhobius	Hidden Markov Model (HMM)	Sec/SPI; Full-protein TM topology	http://phobius.sbc.su.se; [21], [22]
Philius	Dynamic Bayesian Network (DBN)	Sec/SPI; Full-protein TM topology; Protein type	http://www.yeastrc.org/philius; [23]
TOPCONS	Consensus prediction	Sec/SPI; Full-protein TM topology; Protein type	http://topcons.net; [24]
SPOCTOPUS	ANN and HMM	Sec/SPI; TMH	http://octopus.cbr.su.se; [25]
MEMSAT3/ MEMSAT-SVM	ANN; Support Vector Machine (SVM)	Sec/SPI; TMH; Re-entrant helix; Protein type	http://bioinf.cs.ucl.ac.uk/psipred; [26], [27]
DeepSig	Deep Convolutional Neural Network(DCNN); Grammar-Restrained Hidden Conditional Random Field	Sec/SPI; Cleavage site	https://deepsig.biocomp.unibo.it/deepsig; [28]
SigUNet	Convolutional Neural Network (CNN)	Sec/SPI	https://github.com/mbilab/SigUNet; [295]
Signal-3L 3.0	Attention Deep Learning; Window-Based Scoring	Sec/SPI; Cleavage site	http://www.csbio.sjtu.edu.cn/bioinf/Signal-3L; [296]